CALICE Tungsten HCAL Prototype status Erika Garutti Wolfgang Klempt Erik van der Kraaij CERN LCD...

CALICE Tungsten HCAL Prototype status

Erika Garutti Wolfgang Klempt

Erik van der Kraaij CERN LCD

International Workshop on Linear Colliders 2010, October 21

WHCAL experimental setup

IWLC - October 21, 2010 Erik van der Kraaij, CERN LCD 2

CALICE hadronic calorimeter installed at CERN with:

• 30 plates of 1 cm thick tungsten as absorber • 3x3 cm scintillator tiles as active material

with SiPM for readout• Alternative detector technologies will join in

parallel with the normal data taking

Now being commissioned in a muon beam and will move in November to the T9 testbeam line at the CERN-PS (1-10 GeV).

Goal of the test:• Prove the technology of using

tungsten as absorber in an HCAL• Validate hadronic shower

simulations

Installation in PS - T7 area

WHCALmuons1-10 GeV


Three PCs for: - DAQ- Analysis- Power control

Muon flux originating from pions in a PS beamline ~30m away- muon beam size ~1m- Energy 1-10 GeV

Two 50x50 cm2 muon triggers installed

Beam instrumentation

• 2 sets of trigger coincidences: – 10x10 cm2

– 50x50 cm2 (one in front, one behind stack. Not on picture)• 2 delay wire chambers commissioned. 3rd has now also been

installed.IWLC - October 21, 2010 Erik van der Kraaij, CERN LCD

Two delay wire chambers

10x10

4

Beam instrumentation

• One set of 8x1 cm2 to be used for finer measurements• For now not in line with beam, but can easily be moved

IWLC - October 21, 2010 Erik van der Kraaij, CERN LCD

8x1 cm2

5

Delay wire chamber commissioning

• The wire chambers have a delay line running back & forth from left to right, and a line running back & forth from up to down.

• The chamber gives four signals:time delay from the left, right, up and down sides.


LEFT-RIGHT (one chamber)

Tested correlation in time of hits from the left and right signal:• As a particle hit is read out both sides, then the sum of

the two distances (and hence times if the velocity is linear) should be a constant.

Chambers are working correctly

6

Delay wire chamber commissioning


Arrival time for multiple hits on one channel

Red: 1st Blue: 2nd

Black: 3rd

Left channel

Right channel

Down channel

Up channelTime difference between 1st and 2nd

hit [ns]

Calibration has been performed to convert the time signal to position measurements.

• Reconstructing of tracks made difficult by multiple hits.

• Time difference between 1st and 2nd hit peaks at 200 ns; most probably reflection.

Will keep the time windows small to minimize #double hits on one channel

3rd wire chamber is being installed & commissioned at the moment. This will reduce combinatorics.

Muon spills come in ~300 ms window, every 45 sec.

• Relative timing measured and adjusted where needed for triggers and wire chambers.

• Using scintillator triggers:– Coincidence of the two

10x10 cm2 results in ~350 counts/spill

– Coincidence of the two 50x50 cm2 results in~1000 counts/spill

• DAQ Rate: 8 Hz (1 ms deadtime / event)

Triggering and event rate


Spill signal

Muon counter

8

Calorimeter commissioning

Installation done 30 Aug. – 3 Sept.• Noise level consistent with previous installations in

testbeams at CERN (2006/2007) and FNAL (2008)• Large number of dead channels in modules 24/29/30.

(definitions: module = active layer, channel = 1 scintillator tile with SiPM)

• Not yet clear why, the channels sometimes recover.

• These channels were known from last FNAL runs to be malfunctioning

• Yet during pre-assembly at DESY the last few months they showed no problems.

Plans for repair at CERN before November testbeam


Calibration & monitoring system

SiPM response depends on temperature and voltageTemperature sensorsLED monitoring system

One LED illuminates 18 SiPMs and one PIN to monitor the LED signal.

• Functionalities of the LED system: – Gain calibration

(see later)– Provide full dynamic

range for checking the SiPM response function


LED system tested and operational, except for one module (#20). This CMB has been removed for repair.

CMB: Calibration and Monitoring boardT: Temperature sensor

Muon events recorded


In 14 runs collected ~2 million events:

calib. & physics

physics events

events having a track with at

least8 hits in a row

(~60%)

Not all muons come in perpendicular

• Will need to take angle into account in calibrations

Trigger delay scan


GREEN distribution is results from previous testbeam. RED and BLUE are two runs taken this month with the two trigger setups

Delay (induced by trigger setup) is similar for all modules.

• The signal from the SiPM is readout after amplifier & shaper, at its maximum value. Need to determine the delay values (‘hold value’)

• This has been done already at previous testbeams.– Compared with those results the delay values for all modules

should differ by the same amount:

Module number

Tim

e [x

6.25

ns]

calibrated # channels / total # channels:

Gain calibration

Measured amplitude is converted to digital counts: #ADC

• gain = #ADC per pixel

Gain calibration procedure:• use the LED system and take

spectra at low intensity light for all channels

• fit single photon spectra• Gain is difference between 2

single photon peak

Gain values extracted and consistent with previous testbeam

Extraction efficiency in last runs at 95%, almost the level it was for previous testbeam.


efficie

ncy

Run number (module 20 excl.)

error in data acquisition

#ADC counts

MIP energy calibration• Selecting straight tracks, and

applying calibration from 2007, a single channel results in

• Each entry is the energy deposited, expressed in MIP energy as defined in 2007– Small noise distribution– Fit convolution of a Landau

and a Gauss distribution: AMIP = 1.074 ± 0.013 mip


muons

noise

Extracting this average for each channel results in Calibration from 2007 looks reasonable Temperature corrections not yet applied to data

Conversion from signal to energy takes more steps than shown – see talk by Zalesak later.

• Conversion from signal to energy takes more steps than shown – see talk by Zalesak later.

Summary

• Calorimeter installed and operational– LED system operational (apart from 1 CMB)– Repair of 3 modules will be done before moving to T9

• Wire chambers are commissioned • Trigger system timed, DAQ rate ~8Hz.

• Data transfer, conversion & analysis in progress– First results consistent with previous testbeam results– Gain calibration extraction efficiency at 95%

Beginning of November the entire setup will be lifted to the other side of the hall to be placed in PS testbeam.


Backup


Saturation scansOne module (1 m2) consists of 216 scintillator tiles of different

sizes, with for each tile one SiPM measuring the light.• SiPM signal is the sum of signals from fired pixels.

– With limited number of pixels and recovery time (20-500 ns) this result in a non-linear response curve


To extract this response curve, each SiPM is illuminated with increasing intensity

Spread in #ADC saturation approximately factor 2

• As expected from the large variety of SiPM

#ADC

cou

nts

LED Voltage [DAC bin]

Two dead channels

• W-AHCAL will “fly” to T9 on the 3rd of November• After commissioning, official CALICE testbeam starts

on the 6th of Nov.

PS Schedule


W-AHCAL move to T9

possible repair

18

CALICE Tungsten HCAL Prototype status Erika Garutti Wolfgang Klempt Erik van der Kraaij CERN LCD...

Documents

Transcript of CALICE Tungsten HCAL Prototype status Erika Garutti Wolfgang Klempt Erik van der Kraaij CERN LCD...